An AI-Powered Car Towing Management System using Automatic Number Plate Recognition
Sheikh Tareq Ahmed Mentor: Ahmed Ahmed Department of Computer Science Introduction: Automatic Number Plate Recognition (ANPR) is a technology that can recognize vehicle number plates automatically using high-speed cameras. It involves detecting the plat's position in the vehicle, identifying the characters and digits in the plat, and converting the captured image to text. There are many ANPR applications in different domains, such as car park management, traffic management, tolling, and intelligent transport systems. Despite this technology's importance, the existing ANPR approaches suffer from the accurate identification of number plats due to their different size, orientation, and shapes across other regions worldwide. In this project, we tried to study these challenges by implementing an empirical case study for smart car towing management using machine learning models. The developed mobile-based system uses different approaches and techniques to enhance the accuracy of recognizing number plates. Materials and Methods: First, we developed an algorithm to accurately detect the number plate's location on the car body. Then, the bounding box of the plate is extracted and converted into a grayscale image. Second, we applied a series of filters to detect the alphanumeric characters' contours within the grayscale image. Third, the detected the alphanumeric characters' contours are fed into a K-Nearest Neighbors (KNN) model to detect the actual number plat. Results and Discussion: Our model achieves a classification accuracy of 95% in recognizing number plates across different regions worldwide. The GUI has been developed as a form of Android mobile app, allowing law-enforcement personnel to capture a photo of the towed car, which is then recorded in the car towing management system automatically. The app also allows owners to search for their cars, check the case status, and pay fines. Finally, we evaluated our proposed module and the system using various performance metrics such as accuracy, processing time, etc. Conclusion(s): In this project, we developed an ML-powered car towing management system using automatic number plate recognition. We believe that the proposed system would create a better opportunity for law-enforcement personnel to automate car towing. The proposed ANPR model approach comprises three phases: plate detection, character segmentation, and character recognition. We tested our system with an image dataset of various shaped and sized number plates, where crowded backgrounds, low contrast, and diverse illumination condition images are taken into consideration. We carried out several sets of experiments for evaluating the performance and classification accuracy of our system, paying particular attention to the classification and processing time. Our model could most notably process several images per second more than triple the commercial fps. This proves that our ANPR model is suitable for real-time inference at the edge with high prediction accuracy and response time, which could be used for various applications such as tolling, traffic management, security surveillance, etc. Moreover, we found that our model outperforms some state-of-the-art ANPR approaches in terms of the overall processing time. References: [1] M. Y. Arafat, A. S. M. Khairuddin, and R. Paramesran, "Connected component analysis integrated edge based technique for automatic vehicular license plate recognition framework," IET Intelligent Transport Systems, vol. 14, pp. 712–723, 2020. [2] P. Shivakumara, D. Tang, M. Asadzadehkaljahi, T. Lu, U. Pal, and M. Hossein Anisi, "Cnn-rnn based method for license plate recognition," CAAI Transactions on Intelligence Technology, vol. 3, no. 3, pp. 169–175, 2018. 65
